# How to pitch a client with competitor research

> Turn competitor ad research into a short client pitch: what rivals run, where the client is missing, three tests with a budget, and what you cannot know.
- **Author**: Marxx
- **Published**: 2026-11-06
- **Category**: Agencies
- **URL**: https://marxx.ai/library-guides/how-to-pitch-a-client-with-competitor-research

---

To pitch a client with competitor research, show a short and honest picture: what three to five rivals run, which angles and formats they keep paying for, where the client's ads are missing, and the two or three tests you would run first with a budget and a date. Never present guesses as facts.

## Steps

1. **Pick three to five rivals.** Take the client's closest rival, the one they fear and one larger brand.
2. **Collect their live ads.** Use the Meta Ad Library, and the Google Ads Transparency Center for Search and YouTube. For the method, see [how to compare your ads with competitors](/guides/how-to-compare-your-ads-with-competitors).
3. **Tag every ad.** Record the angle, hook, format, offer, start date and number of versions. The [competitor analysis template](/templates/competitor-analysis-template-for-ads) has the columns.
4. **Find the long runners.** Brands often switch off ads that do not pay back, so an ad live for two months or more is a clue, not proof, that it earns its place.
5. **Tag the client's ads the same way,** so the comparison is fair.
6. **Write three findings.** What rivals keep running, what format or angle is missing from the client's ads and what nobody uses.
7. **Turn findings into tests.** Each test gets a guess, a budget, a length and the number that would prove it wrong.
8. **Say what you cannot see.** Ad spend and results are not public for most ads. Run time is a clue, not proof. See [how to tell if a competitor's ad is working](/guides/how-to-tell-if-a-competitor-ad-is-working).

## A five-slide pitch

| Slide | What it shows |
|---|---|
| 1. Where you stand | The client's live ads by angle and format |
| 2. What rivals run | Counts of long-running ads by angle |
| 3. The gaps | Two angles nobody uses and one the client overuses |
| 4. The test plan | Three tests, budget, dates, success numbers |
| 5. What we need | Product access, approvals, a 30-minute weekly call |

## A worked example

Say you pitch a premium tea brand that runs 9 ads. Four rivals run 58 ads, and you tag all of them.

| Angle | Rival ads | Running 60+ days | Client's ads |
|---|---|---|---|
| Health benefits | 14 | 2 | 6 |
| Gifting | 11 | 6 | 1 |
| Taste notes | 9 | 1 | 2 |
| Single-estate story | 3 | 0 | 0 |
| Other | 21 | 3 | 0 |

The client puts 6 of 9 ads on health, where only 2 of the 14 rival ads have lasted 60 days. Rivals keep 6 of their 11 gifting ads live for 60 days or more, and the client has one. Nobody uses the single-estate story. You propose three tests at INR 15,000 each for seven days, INR 45,000 in total: a gifting reel, a taste-notes static and a single-estate story video. You will judge each by cost per purchase against the client's current average.

## Mistakes to avoid

- **Showing rival "spend" numbers.** They are estimates. If you must, give a range and say how you built it. See [how to find competitor ad spend](/guides/how-to-find-competitor-ad-spend).
- **Copying rival ads.** Pitch the angle and your version of it.
- **Counting ads, not ideas.** Ten cuts of one idea are one idea.
- **A 40-slide deck.** Five slides and a spreadsheet are enough.

We make Marxx. It tracks competitor ads on Meta and Google, and its client workspaces come with unlimited seats. See [competitor tracking](/features/competitor-tracking) and [Marxx for agencies](/for/agencies).

## FAQ

### How many competitors should I include?

Three to five. Fewer gives a thin picture. More makes the deck hard to read.

### Can I tell a client how much a rival spends?

No, not from public data for most ads. You can estimate a range from live ad counts and run time, but call it an estimate.

### What if the client is new and has no ads yet?

Pitch from the rivals' side alone. Show what the market runs, then propose a first set of tests. Compare against your own results once you have them.

---
- [More Agencies articles](https://marxx.ai/library-guides/category/agencies)
- [All articles](https://marxx.ai/library-guides)